<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>LazyBuilds</title><description>Product-focused engineer and former startup CTO building scalable systems in AI, Blockchain, and Fintech. Exploring LLMs, RAG systems, and multi-tenant SaaS.</description><link>https://lazybuilds.com/</link><item><title>Belief-based systems: how I learned to stop trusting documents</title><link>https://lazybuilds.com/posts/belief-based-systems/</link><guid isPermaLink="true">https://lazybuilds.com/posts/belief-based-systems/</guid><description>A document is not a fact. It is a claim. After four months of shipping an AI sales rep, I think the next infra layer is belief-based knowledge systems. Here is the buildup, with one running example, six diagrams, and the papers behind it.</description><pubDate>Tue, 05 May 2026 00:00:00 GMT</pubDate></item><item><title>RAG Through Wonderland — Part 2: Finding the Rabbit Hole (Level 1)</title><link>https://lazybuilds.com/posts/part-2-rag-through-wonderland/</link><guid isPermaLink="true">https://lazybuilds.com/posts/part-2-rag-through-wonderland/</guid><description>Building and optimizing a RAG system for Level 1 factual recall—from 12/30 to 27/30 through iterative improvements in chunking, retrieval, and evaluation.</description><pubDate>Mon, 17 Nov 2025 00:00:00 GMT</pubDate></item><item><title>AliceEval - Part 1: Setting Up AliceEval</title><link>https://lazybuilds.com/posts/part-1-rag-through-wonderland/</link><guid isPermaLink="true">https://lazybuilds.com/posts/part-1-rag-through-wonderland/</guid><description>Kick off the RAG Through Wonderland series by defining AliceEval and outlining the five-level journey from factual recall to external knowledge synthesis.</description><pubDate>Mon, 10 Nov 2025 00:00:00 GMT</pubDate></item><item><title>What is Retrieval-Augmented Generation (RAG)?</title><link>https://lazybuilds.com/posts/what-is-rag/</link><guid isPermaLink="true">https://lazybuilds.com/posts/what-is-rag/</guid><description>A practical guide to RAG—what it is, why it matters, benefits, how it works, and trade-offs.</description><pubDate>Wed, 15 Oct 2025 05:17:19 GMT</pubDate></item></channel></rss>